a Non-Mendelian Mutation of Paramecium by Microinjection …We show that microinjection of a 4.5-kb internal A gene fragment is sufficient for proper processing at autogamy and leads

Copyright 0 1991 by the Genetics Society of America

Permanent Rescue of a Non-Mendelian Mutation of Paramecium by Microinjection of Specific DNA Sequences

Helen Jessop-Murray, Linda D. Martin, David Gilley, John R. Preer, Jr., and Barry Polisky Program in Molecular, Cellular, and Developmental Biology, Department of Biology, Indiana University,

Bloomington, Indiana 47405 Manuscript received May 10, 199 1

Accepted for publication August 2, 199 1

ABSTRACT The mutant Paramecium tetraurelia cell line d48 is unable to express the serotype A protein on its

surface. Although the A gene is intact in the micronuclei of d48, the A gene copies in the macronucleus contain a large deletion eliminating virtually the entire coding sequence. Previous studies showed that microinjection of a plasmid containing the entire A gene into the macronucleus of d48 permanently restored A expression after autogamy. Together with other data, this result suggests that in wild type cells the A gene in the old macronucleus ensures the presence of a cytoplasmic factor that prevents A gene deletions at autogamy. In d48, where there are few, if any copies of the intact A gene in the old macronucleus, deletions occur during macronuclear formation. To elucidate the specific molecular mechanisms involved in this unusual phenomenon, we attempted to define the region(s) of the A gene necessary for rescuing d48. We show that microinjection of a 4.5-kb internal A gene fragment is sufficient for proper processing at autogamy and leads to permanent rescue of d48; i.e., the rescued strain is indistinguishable from wild type. Thus, rescue of d48 does not require upstream transcriptional control sequences, intact A mRNA or A serotype protein. We also show that various fragments of the A gene have the ability to rescue d48 to different extents, some being more efficient than others. We find no evidence to suggest that the A gene gives rise to a small stable RNA that might act as or encode a cytoplasmic factor. Molecular mechanisms that may be involved in the rescue of d48 are discussed.

. -

I N wild-type Paramecium tetraurelia stock 51, the A surface protein is coded by the A gene located

close to the telomere in macronuclear chromosomes (reviewed by PREER 1986). In each wild type cell, the polyploid macronuclear chromosomal DNA contains about 2000 copies of the A gene. Macronuclear chromosomes range in size from about 100-600 kb. The micronucleus (containing about 2 100 kb per average chromosome; PREER 1986) is diploid. T h e A gene is present in the micronucleus of a mutant line called d48; however the macronucleus of this mutant contains few, if any copies of the intact A gene (RUDMAN et al . 199 1). As a result, the A protein is not detectably expressed by d48 (EPSTEIN and FORNEY 1984). At autogamy (a self-fertilization process that occurs pe- riodically in P. tetraurelia) and conjugation, a new macronucleus and new micronuclei are formed from DNA processing of the old micronuclei. T h e old macronucleus degenerates as the new macronucleus develops. Although the molecular details of autogamy and conjugation are unknown in Paramecium, the DNA processing steps involved in generating polyploid macronuclear chromosomes from diploid micronuclei include DNA cleavage, telomere addition and DNA amplification. During formation of the macronucleus in the d48 mutant, processing of the A

Genetics 129 727-734 (November, 1991)

surface antigen gene is aberrant. A large chromosomal deletion which begins near the 5’ end of the gene (EPSTEIN and FORNEY 1984; FORNEY and BLACK- BURN 1988) eliminates virtually the whole A gene from the macronucleus.

T h e genetic behavior of the d48 mutation is unusual. A cross of 5 1 (A+) to d48 (A-) usually yields A+ exconjugant F1 clones from the A+ parent, and A- exconjugant clones from the A- parent, despite the fact that both exconjugants are genetically identical. Production of an F2 by the induction of autogamy in the F1 usually yields no further change in the ability to produce the A serotype in the progeny of either exconjugant (EPSTEIN and FORNEY 1984; RUDMAN et a l . 1991). Since one-half of the F2 exconjugants de- scended from the d48 parent are homozygous for the wild type A gene, the inability to observe proper A gene processing implies the existence of factors in an A+ cell responsible for proper processing of the A gene. This non-Mendelian pattern of inheritance shows that the genetic difference between d48 and wild type does not lie in the micronuclei of the two strains.

However, the genetic behavior does not seem to represent true cytoplasmic inheritance either. HARU- MOTO (1986) showed that transfer of macronuclear

728 H. Jessop-Murray et al.

material, but not cytoplasm, from vegetative wild type cells into d48 will "rescue" d48, permanently restoring its ability to produce A after autogamy. KOIZUMI and KOBAYASHI (1 989) found that microinjection of wild- type cytoplasm rescued d48, but only if both donor and recipient were undergoing autogamy. Rescued cells continued to express A even after successive autogamies. We interpret these results to mean that the basis for the d48 mutation is not due to the lack of cytoplasmic hereditary determinants, but instead results from the absence of a cytoplasmic factor that is normally produced during autogamy and conjugation in wild type cells. Its genetic determinant resides within the old macronucleus and is necessary for inclusion of the A gene into the newly forming macronucleus.

Further light was shed on this phenomenon by microinjection of plasmid DNA containing the A gene (GODISKA et al. 1987). When such DNA is injected into the macronucleus of the A- mutant d12, which lacks the A gene in its macronucleus, the plasmid is linearized, acquires Paramecium-type telomeres, and replicates autonomously in the macronucleus (GILLEY et al. 1988). The injected d l 2 cells acquire the ability to transform to A+, but "revert" at autogamy when the old macronucleus is replaced by a new one (GOD- ISKA et al. 1987). T h e elimination of the macronucleus at autogamy results in the physical loss of the injected DNA. These results suggest that the d l 2 mutation involves an alteration of the micronuclear copy of the A gene. Similar transformation to an A+ serotype occurs when the plasmid DNA is injected into the macronucleus of d48 (KOIZUMI and KOBAYASHI 1989), except that in this case the cells remain A+ after autogamy, i e . , d48 is permanently rescued. YOU et al. (1991) have recently shown that a fragment of the A gene is capable of d48 rescue. These results suggest that the d48 phenotype is due to the deficiency of the A gene or sub-A gene sequences in the old macronucleus. Presumably, in wild type cells these sequences in the old macronucleus ensure the presence of a cytoplasmic factor at autogamy and conjugation that prevents loss of the A gene in the newly forming macronucleus. This factor is not present in the cytoplasm of vegetative cells. Once the A gene is lost from the macronucleus in the mutant d48, it is not normally restored at either conjugation or autogamy.

T h e specific molecular mechanisms involved in this unusual phenomenon are of interest. To elucidate these mechanisms we are attempting to define the region(s) of the A gene necessary for rescuing d48. We show that injection of a 4.5-kb internal fragment of the A gene yields proper processing in d48 and leads to permanent rescue to the wild type phenotype. Neither detectable A gene expression nor complete A gene mRNA is required for rescue. We also show that

various fragments have the ability to rescue d48 to different extents, some being more efficient than others.

MATERIALS AND METHODS

Strains and culture conditions: Wild-type cells were P. tetraurelia, stock 5 1. Strain d48 was obtained from stock 5 1 by X-ray mutagenesis and antiserum selection (EPSTEIN and FORNEY 1984). Cells were cultured in 0.15% Cerophyl (Pine Brothers, Kansas City, Missouri) supplemented with 0.1 g/ liter Bacto Yeast extract, 1 mg/liter stigmasterol, 0.45 g/ liter NanHP04, and inoculated with Klebsiella pneumoniae 24-48 hr before use.

For microinjection, autogamy was induced in d48 by starvation. d48 cells were allowed to undergo six fissions at 27" post-autogamy. Samples were also cultured at 34", which induces serotype A expression in stock 5 1 (wild type), and tested with antiserum to confirm that cells had not spontaneously reverted to wild type. After injection of cloned DNA, each injected cell was isolated into a depression slide and cultured for 24 hr at 34", and then a further 24 hr at 27 " (about eight fissions). At this point, a 20O-pl sample from each depression was stored at 14" as a preautogarny stock. The remaining cells were screened for the presence of cloned DNA by either: testing for serotype A expression, or by transferring to 20 ml of medium in a test tube, culturing for 24-48 hr at 27", and analyzing the cellular DNA by dot-blot hybridization (see below). Following either procedure, the lines were then subcultured in tubes at 27" for 22-26 fissions past autogamy. Autogamy was induced by starvation and confirmed by staining. Each line was subcultured for four fissions at 27" postautogamy. To induce serotype expression, fresh medium was added daily to double the culture volume for 4 days at 34" (EPSTEIN and FORNEY 1984), after which lines were scored for antigen A expression.

Microinjection: Microinjection of cloned DNA into the macronuclei of d48 was done essentially as previously described (GODISKA et al. 1987). One difference was that plasmids pSA5.5, pSA2.8, pSA4.5, pSA3.O and pSA2.5 (see below) were injected as cleared lysate preparations (MAN- IATIS, FRITSCH and SAMBROOK 1982). The cloned DNA was dissolved in DRM medium (1 14 mM KC1/20 mM NaC1/3 mM NaH2P04, pH 7.4) at 1 mg/ml.

Cloned DNA: Plasmid pSA 14SB consists of a 14-kb insert (sequences - 1590 to + 1 1825) containing the antigen A gene and a portion of the flanking regions, cloned into the vector pT7/T3-18 (Bethesda Research Laboratories). The sequencing number system used here refers to the A nucleotide of the ATG codon thought to be the first codon of the A gene as + l . Plasmid pSA12.8 (containing sequences -264 to +11825) was constructed from pSA14SB by cleavage at position -1590 with Sal1 and treatment with exonuclease III/mung bean nuclease (MANIATIS, FRITSCH and SAMBROOK 1982). The Klenow fragment of DNA polymerase I was used to fill in the sticky ends and blunt end ligation was carried out according to MANIATIS, FRITSCH and SAMBROOK (1982). Plasmid pSA5.5 (containing sequences +13 to +5482) was made by digesting pSA14SB with XmnI and Ssp1 and isolating the 5.5-kb fragment. BamHI linkers were attached to the blunt ends (MANIATIS, FRITSCH and SAM- BROOK 1982) and the insert was ligated into the BamHI site of pT7/T3-18. Plasmid pSA2.8 (containing sequences +403 to +3175) was made in an identical manner to pSA5.5, except that the insert was a 2.8 kb HgaI fragment whose sticky ends were filled in with Klenow fragment (MANIATIS,

Rescue of a non-Mendelian Mutation 729

- 1 6 9 0 1 1 9 2 5 pSAl4SB - +

- 2 8 4 1 1 8 2 5 pSA12.8 . P I +

1 ¶ 6482 pSA6.6 1 1

Xmrl Srpl

4 0 1 ¶ 1 7 6 pSA2.8 w 7 7 %

I 3 pSA4.6

4617 I 1

Xmrl PSI1

I¶ 2971 pSA3.0

JdZ==@l I

pSAZ.6

+

+

- I +

+ I -

FIGURE 1 .-Map of the cloned A gene and subcloned fragments. The light bar of pSA14SB represents the coding region of the gene, and the dark bars the flanking regions. pSA5.5, pSA2.8, pSA4.5, pSA3.0 and pSA2.5 were constructed from fragments located within the A gene coding region as described in MATERIALS

AND METHODS. The ability of the fragments to rescue d48 by microinjection are indicated by + (rescue), - (no rescue), -/+ (poor rescue) and +/- (intermediate rescue). Stock 51 macronuclear DNA contains the complete A gene, but d48 macronuclear A gene- containing chromosomes lack downstream sequences beyond the vicinity of + 1 , thereby rendering d48 unable to express the A surface antigen.

FRITSCH and SAMBROOK 1982) before attachment to the linkers. Plasmid pSA4.5 (containing sequences +13 to +4517) was an XmnI-Pstl fragment of the A gene ligated into the SmaI-PstI sites of pT7/T3-18. Plasmid pSA3.O (containing sequences + I 3 to +2971) was constructed from an Xmnl-EglII fragment ligated into the filled-in HindIII- BglII sites of pT7/T3-18. Plasmid pSA2.5 (containing sequences +2971 to +5482) used the same vector sites as pSA3.0, except that the A gene fragment was a BglII-Ssp1 digest. Maps of all these plasmids are shown in Figure 1.

Screening preautogamy injected d48 cells for cloned DNA: Screening was performed on preautogamy injected lines that had been transferred from depressions to 20 ml of medium in test tubes and subcultured to 1000 animals/ ml (cells were 16-20 fissions past the previous autogamy). Either of two methods was used to extract and analyze DNA: (1) 19,000 animals (3 pg DNA) were resuspended in 0.02 ml of their own culture fluid and added quickly to 0.04 ml NDS medium (1 % SDS/O.5 M Na2-EDTA/lO mM Tris- HCI, pH 9.5) at 65" for 24-48 hr. Lysates could be stored at 4". One microliter of 3 M NaOH was added to 10 pl of the NDS mixture, and after heating to 65" for 30 min to denature the DNA, samples were rapidly cooled on ice and neutralized with 11 pl of 2 M NH40Ac. Duplicate 1-pl samples (20 ng DNA) were spotted onto a piece of dry nitrocellulose. (2) Fifty microliters of 3 M NaOH were mixed with 500 PI (500 animals/75 ng DNA) of each culture and heated to 65 " for 30 min. After cooling on ice and neutral- izing with 37.5 p1 of 7.5 M NH40Ac, each sample was transferred to nitrocellulose using a Schleicher and Schuell minifold I system.

Paramecium DNA preparation: NDS lysates of pre- and postautogamous injected lines were prepared as described by GODISKA et al. (1987). DNA was purified from 300 pl of the NDS mixtures by adding 200 p1 of water, extracting

with 500 pl of phenol, reextracting the phenol phase with 500 pl of TE (10 mM Tris-HCI/l mM EDTA, pH 8.0), treating with Sevag and precipitating with 2 volumes of ethanol for 10 min in an ice bath. After washing the precip itate with 75% ethanol and desiccating, the DNA was resuspended in TE.

Paramecium DNA blots: DNA (5 pg) was Cut with HindIII, separated on a 0.8% agarose gel and transferred onto nitrocellulose as described by MANIATIS, FRITSCH and SAMBROOK (1982).

For dot-blots, postautogamy DNA (300 ng) was dena- tured with 0.1 volume of 3 M NaOH at 65" for 30 min, neutralized with 7.5 M NH40Ac and transferred onto nitrocellulose using a Schleicher and Schuell Minifold I system.

Nick translation: DNA was nick-translated to produce probes with specific activities of 5 X 10' to 1 X 10' cpm/pg by procedures described in MANIATIS, FRITSCH and SAM- BROOK (1982).

DNA blot hybridization: Prehybridization at 42" was in 50% formamide, 5 X SSC, 1 X Denhardt's solution, 50 mM Hepes (pH 7.0), 1 mM EDTA, and 100 pg/ml salmon sperm DNA. Hybridization was in the same buffer containing 10% dextran sulfate 5000. Filters were washed twice with 2 X SSC/O.l% SDS for 10 min at 23", and either once or twice with 0.1 X SSC/O. 1% SDS for 20-30 min at 68" prior to autoradiography.

Paramecium RNA preparation: Whole cell RNA was extracted by lysing cells in guanidine hydrochloride as previously described (PREER, PREER and RUDMAN 1981).

Paramecium RNA blots: Whole cell RNA (25 pg) and RNA markers (BRL) were separated on a 1% agarose- formaldehyde gel. After soaking the gel in 20 X SSC and staining with ethidium bromide to confirm equal sample loading, RNA was transferred onto nitrocellulose (MANIA- TIS, FRITSCH and SAMBROOK 1982). In vitro synthesis of RNA hybridization probes: ["PI

UTP-labeled RNA complementary to each strand of plasmid pSA2.5 was transcribed in vitro using conditions similar to those described by MELTON et al. (1 984). One microgram of linearized DNA templates was transcribed in a volume of 20 p1 containing 40 mM Tris, pH 8.0, 8 mM MgC12, 2 mM spermidine, 25 mM NaCI, 10 mM dithiothreitol, 500 p~ of ATP, CTP and GTP, 10 p~ UTP, 7.5 p~ [52P]UTP (800 Ci/mmol), and 13 units of T7 RNA polymerase or 10 units of T 3 RNA polymerase. Synthesis was at 37" for 60 min. The mixtures were then treated with 15 units of RNAse- free DNAse (Pharmacia) for 10 min at 37", phenol-extracted, and the RNA ethanol precipitated with carrier tRNA. Each probe had a specific activity of about 5 X 10'

RNA blot hybridization: Prehybridization was in 50% formamide, 5 X SSPE, 0.1% SDS, 5 X Denhardt's solution, 1 mM EDTA, 200 pg/ml salmon sperm DNA and 100 pg/ ml tRNA at 55". Hybridization was in the same buffer containing 10% dextran sulphate 5000 and 2.5 X Den- hardt's solution. The nitrocellulose was washed twice with 2 X SSC/O. 1% SDS, once with 1 X SSC/O.l% SDS, and once with 0.1 X SSC/O.l% SDS, each at 65" for 20 min.

Primer extension: Primer extension analysis was carried out essentially as described by SAMBROOK, FRITSCH and MANIATIS (1 989). Whole cell RNA (1 0 pg) was hybridized with a 20 nucleotide primer complementary to A gene mRNA sequences from positions +62 to +43. The primer extension reaction was then electrophoresed in a 6% poly- acrylamide/8 M urea gel and autoradiographed. The labeled products were compared to chain termination sequencing reactions using the above mentioned primer to precisely determine the extension product size.

cpm/pg.

730 H. JessopMurray et al.

TABLE 1

Microinjection of plasmid pSA12.8

Preautogamy Postautogamy

Total lines Total lines DNA Injected tested Serotype tested Serotype

pSA12.8 (-264 to 11825) 68 12 A+ 12 1 1 A+ 56 A- 2 4 6 A +

18 A- Uninjected 48 48A- 12 12A-

RESULTS

Injection of pSA12.8: Previous workers have shown that microinjection of plasmid pSA14SB into the macronucleus of d48 causes the cells to express serotype A both before and after the next autogamy (KOIZUMI and KOBAYASHI 1989). To extend these results we microinjected an upstream deletion derivative of pSA14SB, plasmid pSA12.8 (see Figure l), to observe whether it could also induce the mutant cells to express serotype A. pSA12.8 has 264 bp upstream of the presumed ATG start of the A protein while pSA 14SB has 1590 bp. Previous work in our laboratory has shown that pSA12.8 contains the minimal upstream sequences necessary for a fully functional promoter, as well as the entire coding region of the A gene (L. D. MARTIN, unpublished data).

Throughout this report, we use two terms to de- scribe cells that have been successfully microinjected with plasmid DNA. Prior to autogamy, injected animals expressing serotype A and/or shown to contain autonomously replicating cloned A genes or fragments of A genes in their macronucleus are said to be TRANSFORMED. After autogamy, the cloned DNA is lost and cells subsequently able to express serotype A due to the presence of macronuclear copies of the A gene, are said to be RESCUED.

The results of microinjecting supercoiled pSA12.8 into the macronucleus of d48 are shown in Table 1. Preautogamy transformation was scored initially by testing for serotype A expression. Of 68 injected cells, 12 gave rise to transformed lines. Twelve A+ lines and 24 of 56 A- lines were serotype tested every day during subculturing for a further 12- 16 fissions (see MATERIALS AND METHODS). It was found that 7 of the 12 A+ lines lost the ability to express A prior to autogamy. However, once through autogamy 11 of the 12 transformed lines were rescued, and surpris- ingly, so were 6 of the 24 preautogamy A- lines. Preautogamy stocks at 14" of these unusual lines were expanded and whole cell DNA was extracted to determine the presence of the injected plasmid. Unin- jected d48 acted as a control in the microinjection experiment; no spontaneous reversion to the wild type phenotype was observed on subculturing the cells through autogamy (Table 1).

Pre-autogamy serotype

Q)

4 n 0 z :

F c ' D o 0 a + I o z s y m 4 4 a a a 8 - 0 P

1 2 3 4 5 6 7 8 .

= 3:: 2.3 2.2 - 1.4

I 0.8 0.7 - 0.5

FIGURE 2.-Hybridization of labeled pSA12.8 DNA to HindIII digested whole cell DNA (5 pg) isolated from preautogamy d48 animals injected with pSAl2.8. All lines were rescued postautogamy to serotype A. Lanes 1 and 2: lines that showed stable serotype A expression preautogamy. Lane 3: a line initially expressing serotype A that lost expression immediately prior to autogamy. Lanes 4 and 5: lines that did not express serotype A preautogamy. Lane 6: wild- type DNA. Lane 7: uninjected d48. Lane 8: pSA 12.8 digested with HindIII. Lanes 1 to 6 were shown to have equal loading of DNA by ethidium bromide staining of the gel. Lane 7 contained less material. Numbers to the right of the blot refer to the sizes in kb of the HindIII fragments of pSA12.8 DNA which sewed as size markers.

In all cases, lines rescued postautogamy were shown to contain plasmid pSA12.8 DNA preautogamy. As described above, some of the transformed lines did not express A surface antigen as determined by the serotype assay. DNA (5 pg) from several such preautogamous A+ and A- lines was cleaved with Hind111 and probed with labeled pSA12.8 DNA after South- ern blotting (Figure 2). The restriction pattern obtained from transformed cell DNA (lanes 1-5) was identical to that of the cleaved plasmid pSA12.8 DNA (lane 8), rather than to that of wild-type macronuclear A genes (lane 6). The results show that in lines unable to express serotype A pre-autogamy (lanes 4 and 5), there are fewer copies of plasmid DNA present than in transformed A expressing cells (lanes 1 and 2). This low copy number is presumably insufficient to allow for serotype expression pre-autogamy, but sufficient for rescue to occur postautogamy. Lane 3 shows DNA from a line that initially was A+, but lost expression prior to autogamy. The DNA sample was obtained from animals that were 20-22 fissions postinjection when A expression was lost. However, plasmid DNA was present in this line at a copy number similar to that in lines unable to express serotype A preautogamy (lanes 4 and 5). It is possible that more plasmid DNA

Rescue of a non-Mendelian Mutation 73 1

was present initially, allowing for A expression, and that a decrease in plasmid copy number occurred during subsequent subculturing. One important conclusion that can be made from these unusual lines is that the injected plasmid, but not high-level A expression prior to autogamy, is required for rescue.

Analysis of the DNA from postautogamous rescued cell lines showed that wild-type macronuclear chromosomes containing the A serotype gene were present, rather than cloned A genes (results not shown). DNA requirements for d48 rescue: We cloned a

variety of restriction fragments of the A gene into the plasmid vector pT7/T3-18 to observe whether specific fragments were capable of rescue. Unlike pSA12.8, these subclones lacked transcriptional and translational information required for A gene expression and therefore could not transform d48 to express serotype A. The structure of the subclones is shown in Figure 1.

Plasmid DNA was cleaved with restriction enzymes to free the A gene sequence containing inserts from the vectors before injection. This was done to ensure that the orientation of the fragments within the vector would not interfere with their ability to rescue d48. Because preautogamy transformation could not be scored by A expression, the DNA of each injected line was analysed by dot-blot hybridization. The blots were probed with nick-translated fragments of the A gene to assess the presence of plasmid, and corrected for loading errors using a probe for the Paramecium a- tubulin gene. Rescue post-autogamy was measured by the ability of the cells to express serotype A, as before.

In one experiment, plasmids pSA5.5 and pSA2.8 (see Figure 1) were compared for their ability to rescue d48. pSA5.5 contains a fragment of the A gene from +13 to +5482 of the coding region. The cloned fragment in pSA2.8 carries +403 to +3175 of the coding region. The results of the injections are shown in Table 2. Both plasmids were efficiently established in 20 out of about 100 injected cells. The copy number of each plasmid varied within the 20 lines, but the range of copy numbers was similar (results not shown). Any differences between post-autogamous rescue efficiency was therefore not due to one plasmid being more efficiently established in the preautogamous cells than the other. The postautogamy serotype pa- rameter “% A+” represents the result of the serotype assay, a score of 100 meaning that all the test animals were immobilized with antiserum, 40 meaning that only 40% of the test population were immobilized, and so on. Any line exhibiting a value between 1 and 40% A+ was designated a “mixed culture” (HARU- MOT0 1986; RUDMAN et d . 1991). We observed that if animals were isolated from these postautogamy pop- ulations and subcultured, some of the secondary lines were capable of expressing A and others were not

(data not shown). It appears that in these mixed cultures only a fraction of the progeny of one injected cell was rescued and acquired the ability to produce normal macronuclei. We do not know why this occurs, but presumably rescue in any individual animal will depend upon the plasmid copy number in that cell at autogamy and/or the efficiency of rescue of the A gene fragment itself, mixed cultures resulting when either one is low.

The data in Table 2 show that 19 out of the 20 lines transformed with plasmid pSA5.5 were rescued, with a majority of the postautogamy lines showing between 40 to 100% A+. Of the remaining 80 lines, three showed a low percentage of A expression and this probably represents spontaneous reversion of the mutant d48 cells to wild type, or alternatively transformation with undetectable levels of plasmid. In com- parison, only one of the 20 lines transformed with plasmid pSA2.8 was rescued to A expression, and none of the remaining lines were able to express A. Uninjected controls showed no reversion to wild-type cells. The conclusion that can be drawn from the injection of pSA5.5 and pSA2.8 is that specific A gene sequences found in the 5.5-kb fragment are responsible for efficient d48 rescue and that neither A expression nor intact mRNA preautogamy is required for rescue.

To characterize the sequence requirement further, three other subclones were injected. The fragment in plasmid pSA4.5 has 1 kb downstream removed compared to the 5.5-kb fragment, and contains +13 to +45 17 of the coding region (see Figure 1). Plasmids pSA3.O and pSA2.5 are subclones of the 5.5-kb fragment cleaved at +2971. The upstream portion, pSA3.0, consists of +13 to +2971 of the coding region, and the downstream portion, pSA2.5, contains +2971 to +5482. The results of the injections are shown in Table 2. All three plasmids were established in injected cells with similar efficiency preautogamy (data not shown). Post-autogamy, pSA4.5 was able to rescue d48 efficiently, with most of the rescued lines showing a high percentage of A expression (between 40 and 100% A+). pSA3.O rescued the mutant with poor efficiency, a majority of the lines giving a weak postautogamy serotype result of 20% A+ and lower. On the other hand, pSA2.5 gave an “intermediate” type of rescue reaction with half of the postautogamy lines expressing between 1 and 20% A+, and the other half expressing between 21 and 60% A+. With all three plasmids post-autogamy rescue correlated with the presence of injected DNA pre-autogamy. These results indicate that specific sequences of the A gene have the ability to rescue d48 to different extents, some being more efficient than others.

When the results of microinjecting plasmids pSA5.5 and pSA4.5 are compared (Table 2), it can be seen


TABLE 2

Frequency of transformation and rescue from microinjection of cloned A gene fragments

Postautogamy Preautogamy

Serotype (% A+)

Experiment No. DNA lniected tested Plasmid 1-20 21-40 41-60 61-80 81-100 %Rescue

Total lines

~

la pSA5.5 ( I 3 to 5482) 100 20+ 3 3 3 3 7 19/20 = 95 80- 3 0 0 0 0 3/80 = 4

82- 0 0 0 0 0 0/82 = 0

I C Uninjected 32 32- 0 0 0 0 0 0/32 = 0

31- 0 0 0 0 0 0/31 = 0

Ib pSA2.8 (403 to 3 175) I02 20+ 1 0 0 0 0 1/20 = 5

2a pSA4.5 (1 3 to 45 17) 54 23+ 1 1 6 9 6 23/23 = 100

2b pSA3.O (1 3 to 297 1) 63 17+ 16 1 0 0 0 17/17 = 100 46- 1 0 0 0 0 1/46 = 2

2c pSA2.5 (2971 to 5482) 56 24+ 13 6 5 0 0 24/24 = 100 32- 0 1 0 0 0 1/32 = 3

2d Uninjected 49 49- 0 0 0 0 0 0/49 = 0

that the two are essentially indistinguishable in their ability to efficiently rescue d48. Plasmid pSA2.5 gives an intermediate extent of rescue, pSA3.0 a low but detectable rescue, and pSA2.8 is unable to rescue the mutant.

One concern was that the serotype assay did not reflect the molecular status of the rescued lines, and that even though the cultures were expressing low percentages of A, the cells actually had a wild-type copy number of A genes in their macronuclei. To address this concern, purified DNA from a random selection of post-autogamous animals from each serotype group and from each injection experiment was analysed by dot-blot hybridization. Duplicate blots were probed with the A gene and with a probe specific for the Paramecium a-tubulin gene. The A gene- specific hybridization signal was normalized to that of a-tubulin to calculate the A gene copy number. The relationship between the extent of the serotype reaction and the A gene copy number is shown in Figure 3. A positive correlation is observed between A gene copy number and the serotype reaction. The data underscore the fact that pSA5.5 and pSA4.5 can rescue d48 with high efficiency, while pSA2.5 rescues with “intermediate” efficiency and pSA3.O with poor efficiency.

It is unknown why lines rescued by injection of plasmids pSA4.5, pSA2.5 and pSA5.5 expressed such a wide range of postautogamy serotype levels. We examined the possibility that this range in serotype levels was due to variation in the preautogamy copy number of the injected plasmid among the different lines. No such correlation was observed when dot- blots containing pre- and post-autogamy DNA from

FIGURE 3.-Relationship between copy number of A genes and extent of serotype expression in randomly selected postautogamy lines. The lines were generated from individual cells injected with the five plasmids shown to the right of the graph. The copy number of A genes present in the various lines was determined by DNA blot hybridization.

lines injected with plasmids pSA4.5, pSA3.O and pSA2.5 were analysed (results not shown). Thus, the pre-autogamy plasmid copy number in a population is not linearly related to the extent of rescue observed in the postautogamy population. Whether a linear relationship exists at the cellular level between copy number of injected plasmid and post-autogamy serotype response requires analysis of individual cells carried through autogamy.

RNA analysis: KOIZUMI and KOBAYASHI (1989) concluded that a factor responsible for inclusion of the A gene into the newly forming macronucleus is present in the cytoplasm of autogamous and conju- gating wild-type cells. We examined the possibility that a stable RNA from the A gene region might play a role in this process.

Whole cell RNA was extracted from wild-type cells

Rescue of a non-Mendelian Mutation 733

1 fiss 2 fiss past past

log autog autog autog ""

co co w

co d

co d d

~ ~ u a o u ~ n u a n u

1 2 3 4 5 6 7 8 9101112

= 9.5

" 7 . 5

- 4.4

= 2.4

= 1.5

- 0.2

FIGURE 4.-Hybridi~ation of an RNA probe complementary to a portion of the A gene mRNA with Paramecium whole cell RNA. The labeled probe was synthesized by in vitro transcription of plasmid pSA2.5 as described in MATERIALS AND METHODS. Whole cell RNA was extracted from wild-type animals expressing serotype A (lanes 1 , 4, 7, lo), serotype D (lanes 2, 5, 8, 11). and from d48 cells (lanes 3 , 6 , 9 , 12). RNA preparations were made from animals in various growth and developmental stages including log phase (lanes 1 , 2, 3), autogamy (lanes 4, 5, 6), one fission past autogamy (lanes 7, 8, 9). and two fissions past autogamy (lanes 10, 1 1 , 12). The Northern blot was exposed to X-ray film for 10 min at room temperature. Numbers to the right of the blot refer to the sizes in kb of marker RNAs (BRL).

expressing serotype A and D, and from d48 cells. RNA preparations were made from cells in log phase, autogamy, one fission past autogamy and two fissions past autogamy. The RNA was electrophoresed, blot- ted onto nitrocellulose and separately hybridized with two labeled RNA probes complementary to the strands of plasmid pSA2.5. The RNA probes were generated by T7/T3 RNA polymerase transcription in vitro as described in MATERIALS AND METHODS.

The probe complementary to the A gene mRNA hybridized to two RNA species isolated from wild- type cells expressing the A gene (Figure 4, lanes 1, 4, 7, 10). These two RNA species were present in a p proximately the same relative abundance in RNA preparations from wild-type cells at different growth and developmental periods. One species was 8.5 kb in length, which is identical to the size of A gene mRNA reported by PREER, PREER and RUDMAN (1981). A second hybridizing species was roughly 10 kb. The probe did not hybridize with RNA from cells expressing serotype D after short exposure times, or to RNA

prepared from d48 (Figure 4, lanes 2, 3, 5, 6, 8, 9, 11, 12). At longer exposures both A gene mRNA species were faintly detectable in cells stably expressing serotype D (data not shown). Primer extension analysis using RNA isolated from A-gene expressing cells indicated a single 5' end of the A message located seven bases upstream from the ATG presumed to be the initiation codon (results not shown). It is possible that the IO-kb species represents a different con- former of the 8.5-kb species, or alternatively is a result of heterogeneity in transcriptional termination. No transcripts smaller than full-length mRNA that could be candidates for the factor were detected with this highly sensitive probe in any of the RNA preparations (Figure 4), even when longer autoradiographic exposures were examined. When the RNA probe that is not complementary to the A gene message was hybridized to the blot, no signal was observed in any of the whole cell RNA extracts (results not shown), demon- strating that no stable "antisense" RNA is present. We were therefore unable to detect a stable RNA cytoplasmic factor.

DISCUSSION

The non-Mendelian pattern of inheritance shown by d48 represents a macronuclear deficiency that prevents proper processing of micronuclear A genes into the newly forming macronucleus at autogamy. Transformation and subsequent rescue by microinjection of a plasmid containing the A gene into the macronucleus of d48 shows that the d48 phenotype is related to the absence of the A gene in the old macronucleus (KOIZUMI and KOBAYASHI 1989). KOIZUMI and KOBAYASHI (1 989) also showed that a cytoplasmic factor produced during autogamy was responsible for proper inclusion of the A gene into the newly forming macronucleus. Taken together, these results suggest that in wild-type cells, the A gene in the old macronucleus ensures the presence of a cytoplasmic factor that is responsible for the proper processing of the A gene at autogamy. In d48 cells, where there are few, if any intact copies of the A gene in the old macronucleus (EPSTEIN and FORNEY 1984), deletions occur in the newly forming macronucleus at autogamy.

We have defined a region of the A gene that is sufficient to rescue d48. A 4.5-kb internal fragment of the A gene, from +13 to +45 17 of the coding region rescues d48 as efficiently as a plasmid carrying the intact gene. Thus, rescue of d48 does not require upstream transcriptional control sequences, intact A mRNA or A serotype protein. An unexpected result was the "intermediate" extent of rescue obtained with the 2.5-kb downstream A gene fragment (from +297 1 to +5482 of the coding region). Combining the results of microinjection of the 4.5- and 2.5-kb fragments indicates that the region +297 1 to +4517 of the A


gene contains critical sequences. However, the 3.0-kb “upstream” fragment (from +13 to +2971 of the coding region) gave a low but detectable level of rescue. Comparing this fragment to the 2.8-kb portion (from +403 to +3175) which was not capable of rescuing d48, suggests that the region +13 to +403 may also contribute to the rescue process.

Recently, YOU et al. (1991) published a report rel- evant to work reported here. They showed that microinjection of an 8.8-kb fragment internal to the A gene rescued d48. The 4.5-kb fragment reported here is internal to the 8.8-kb fragment.

The molecular mechanism for the involvement of a region of the A gene in the old macronucleus with DNA processing of the micronuclear A genes is unknown. We found no evidence for a small, stable RNA that could either represent or encode a cytoplasmic factor. However, we cannot rule out the possibility that a short-lived, unstable RNA may be synthesized at autogamy.

Alternatively, the injected DNA itself might act directly as the factor. At autogamy, when the macronucleus is degenerating, fragments of the A genes may be liberated into the cytoplasm. These could act in a variety of ways. For instance, they may sequester a processing factor that if left unbound cleaves A genes in the newly forming macronucleus upstream of the coding region. Some liberated A gene fragments may have a greater affinity for the processing factor than others, thereby sequestering it more efficiently. This could explain why the cloned A gene fragments injected into d48 either showed full, intermediate, or very poor rescue. Alternatively, the released A genes may act directly on DNA processing in a yet unknown manner.

The unusual type of genetic behavior manifested by d48 is not restricted to inheritance of the A gene. Other traits in ciliates show similar patterns of inheritance, such as mating type and trichocyst mutants in P. tetruureliu (SONNEBORN 1975; SONNEBORN and SCHNELLER 1979), and surface proteins in Tetruhy- mena thermofihilia (DOERDER and BERKOWITZ 1987). In these three cases, the old macronucleus is involved in the passage of information about its molecular status to the newly developing macronucleus at autogamy and conjugation. The presence or absence of specific DNA sequences in the old macronucleus may be responsible for controlling these traits.

We thank TIM FITZWATER for his technical assistance with the in vitro synthesis of RNA probes. The research reported here was supported by National Institutes of Health grant GM 31745-08.

LITERATURE CITED

DOERDm, F. P., and M. S. BERKOWITZ, 1987 Nucleocytoplasmic interaction during macronuclear differentiation in ciliate pro- tists: genetic basis for cytoplasmic control of SerH expression during macronuclear development in Tetrahymena thermophilia. Genetics 117: 13-23.

EPSTEIN, L. N., and J. D. FORNEY, 1984 Mendelian and non- mendelian mutations affecting surface antigen expression in Paramecium tetraurelia. Mol. Cell. Biol. 4 1583-1590.

FORNEY, J. D., and E. H. BLACKBURN, 1988 Developmentally controlled telomere addition in wild-type and mutant Parame- cia. Mol. Cell. Biol. 8: 251-258.

GILLEY, D., J. R. PREER JR., K. J. AUFDERHEIDE and B. POLISKY, 1988 Autonomous replication and addition of telomerelike sequences to DNA microinjected into Paramecium tetraurelia macronuclei. Mol. Cell. Biol. 8: 4765-4772.

GODISKA, R., K. J. AUFDERHEIDE, D. GILLEY, P. HENDRIE, T. FITZWATER, L. B. PREER, B. POLISKY and J. R. PREER, JR., 1987 Transformation of Paramecium by microinjection of a cloned serotype gene. Proc. Natl. Acad. Sci. USA 8 4 7590- 7594.

HARUMOTO, T., 1986 Induced change in a non-Mendelian determinant by transplantation of macronucleoplasm in Paramecium tetraurelia. Mol. Cell. Biol. 6 3498-3501.

KOIZUMI, S., and S. KOBAYASHI, 1989 Microinjection of plasmid DNA encoding the A surface antigen of Paramecium tetraurelia restores the ability to regenerate a wild-type macronucleus. Mol. Cell. Biol. 9: 4398-4401.

MANIATIS, T., E. F. FRITSCH and J. SAMBROOK, 1982 Molecular Cloning, A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

MELTON, D. A., P. A. KRIEG, M. R. REBAGLIATI, T. MANIATIS, K. ZINN and M. R. GREEN, 1984 Effxcient in vitro synthesis of biologically active RNA and RNA hybridization probes from plasmids containing a bacteriophage SP6 promoter. Nucleic Acids Res. 12: 7035-7056.

PREER, J. R., JR., 1986 Surface antigens in Paramecium, pp. 301- 339 in Molecular Biology of the Ciliated Protozoa, edited by J. G. GALL. Academic Press, New York.

PREER, J. R., JR., L. B. PREER and B. M. RUDMAN, 1981 mRNAs for the immobilization antigens of Paramecium. Proc. Natl. Acad. Sci. USA 78: 6776-6778.

RUDMAN, B., L. B. PREER, B. POLISKY and J. R. PREER, JR., 1991 Mutants affecting processing of DNA in macronuclear development in Paramecium. Genetics 129 47-56.

SAMBROOK, J., E. F. FRISCH and T. MANIATIS, 1989 Molecular Cloning, A Laboratosy Manual, Ed. 2. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

SONNEBORN, T. M., 1975 Paramecium tetraurelia, pp. 469-594 in Handbook of Genetics, Vol. 2, edited by R. KING. Plenum, New York, N.Y.

SONNEBORN, T. M., and M. V. SCHNELLER, 1979 A genetic system for alternative stable characteristics in genomically identical homozygous clones. Dev. Genet. 1: 21-46.

You, Y., K. AUFDERHEIDE, J. MORAND, K. RODKEY and J. FORNEY, 1991 Macronuclear transformation with specific DNA fragments controls the content of the new macronuclear genome in Paramecium tetraurelia. Mol. Cell. Biol. 11: 1133-1 137.

Communicating editor: S. L. ALLEN

a Non-Mendelian Mutation of Paramecium by Microinjection …We show that microinjection of a 4.5-kb internal A gene fragment is sufficient for proper processing at autogamy and leads

Documents